1. Field of the Invention
This invention relates to a video signal recording apparatus for recording video signals and its reproducing apparatus and more particularly, to a video signal recording apparatus and reproducing apparatus capable of obtaining a picture having high picture quality in a high speed reproduction mode operating at a speed which is as high as n times a normal speed and suppressing a degradation in picture quality even if an erroneous MPEG (Moving Picture Coding Expert Group) coded signal intra-/inter-picture adaptively coded digital video signal for ATV (Advanced Television) now under consideration in America was digitally recorded in a VCR (Video Cassette Recorder).
2. Description of the Prior Art
Such a video signal recording apparatus has been conventionally proposed that when recording a video signal on a recording medium, by recording newly formed data for the high speed reproduction use into a sub-code area which is a position corresponding to a trace pattern of a head when reproducing at high speed, the picture quality in high speed reproduction can be improved (refer to Japanese Laid-Open Patent Application No. 2-94071).
With the conventional apparatus, it is presupposed that an input video signal is subjected to coding so as to be entirely finished on an intra-picture basis and yet, the amount of thus coded data is constant for each picture.
With the conventional apparatus, however, such a problem has arisen that an improvement is limited to a picture that is to be reproduced at a speed of two times the normal speed.
With the conventional apparatus, in addition, there arises such a problem that it cannot be applied for a case that the input video signal is coded by inter-picture coding or it is not coded with a fixed length code for each picture. As a result, intra-/inter-picture adaptively coded input video signal data, as with MPEG or ATV, does not have a constant data amount for each picture and the inter-picture coded part thereof cannot be directly decoded, so that the input video signal cannot be recorded.
Further, when considering upon a the fact that valiable length data for the high speed reproduction use is recorded into the sub-code area, it is necessary to ensure a range where the maximum value is considered, and thus a problem arises in which an unnecessary range may occur.
Also, an error correction code generally uses a product code that performs correction in a track unit. As a result, if the reproduction is made so that the head obliquely traverses recording tracks during the high speed reproduction, the inner code of the product code can be employed but an outer code thereof cannot, and a problem arises in which the error correction ability will be degraded compared with that of normal reproduction.
Furthermore, even with respect to the inner code of the product code, if the input signal does not become a fixed length in a macro-block unit, in a case that a packeting is made simply in the order of data transmission, the picture quality during the high speed reproduction is disadvantageously reduced if erroneous. Namely, since the coded data have variable length in the macro-block unit, the decoding can be correctly achieved only in a case that data showing the beginning or finishing of each macro-block can be reproduced sequentially from the beginning, so that in a case that packets cannot be sequentially reproduced due to error occurrence or high speed reproduction, the picture quality will be largely degraded.
In order to solve the above-mentioned problems, for example, a proposal entitled "Video Signal Coding Apparatus" is disclosed in Japanese Laid-Open Patent Publication No. 4-220889. However, even when a top position of a block coded signal can be detected and a coding mode can be decided, in the case of the MPEG signal, necessary information for decoding, that is, a direct current component of a coefficient in a case of intra-picture coding mode and a motion vector information in a case of inter-picture coding mode, are differentially coded and transmitted, so that the block cannot be correctly decoded unless a preceding micro-block can be correctly decoded.